Snapper+GWRC+NZTA. Why integrated ticketing isn’t as simple as it should be.

By Steve Biddle, in , posted: 3-Dec-2015 20:37

Stuff today reported what was almost an obituary for Snapper. While it may have come as a surprise to some, it was hardly news. The real story is the one behind the scenes.

Before we look into the bigger issue, lets take a look at the history of Snapper. Infratil subsidiary NZ Bus own and operate bus services in Wellington, Lower Hutt and Upper Hutt under the Go Wellington, Valley Flyer and Runciman’s brands under contract. By 2006 their existing card based ticketing solution was horrible and cash fares were a nightmare to deal with. They needed a solution, and that solution was a contactless smartcard. Infratil also had just the man to launch a product, and with that, Snapper was born. The company was headed up by Charles Monheim. Monheim, who had headed up Tfl (Transport for London’s) Oyster Card ticketing system from 2001 to 2006 had the perfect skillset required to launch a new ticketing system.

Snapper opted for a system based on the Korean T-money platform, with the actual Snapper “card” bring a JCOP (Java Card Open Plaform) application that resides on the card. This platform had been in use in Korea for a number of years and had a number of advantages over other solutions, particularly competing MIFARE based solutions at the time that were dealing with ongoing issues with the encryption on cards being broken rendering the cards open to being compromised. 

Snapper hit the ground running in July 2008 – literally. Problems arose with buses that didn’t yet have readers, and the accuracy of billing for some journeys was a nightmare that took quite some time to solve. Over the coming months however the issues were sorted, and as we now look at the product 7 1/2 years on from the launch it’s safe to say it’s fulfilled it’s purpose of delivering a solid contactless solution for Wellington bus customers.

Within months of it’s launch in 2008 it was clear Snapper were out to play hardball. Attempts to convince both Mana and Newlands bus companies to offer the product failed – in part because both took issue with the link between Snapper’s owner Infratil, and NZ Bus. As Snapper would have access to full details of passenger numbers and routes the fear was they would then provide this to NZ Bus who could potentially then tender against Newlands or Mana to operate services. This lead to the GWRC (Greater Wellington Regional Council) also privately having concerns about expansion of Snapper – what would happen if NZ Bus lost the tender for bus services but controlled all the ticketing? What conflicts of interest are there when both share the same parent? They’re all issues that have plagued Snapper since the launch, and to an extent still plague it today.

2008 also marked the announcement by ARTA (Auckland Regional Transport Authority) of it’s plans for a new integrated ticketing solution for Auckland across all buses, trains, and ferries. NZ Bus owned several bus companies in Auckland so took the initiative to launch Snapper into the Auckland market, promising that they could deliver a full solution and have it in place before the 2011 Rugby World Cup at a fraction of the cost of other suppliers. Unfortunately for Snapper several key members of the ARTA had some well known personal grievances with NZ Bus, which in turn meant Snapper was on the out, before it was even in.

It was around this time that the NZTA (New Zealand Transport Agency) stepped into play pledging to help fund integrated ticketing in Auckland, providing they had a say in the solution and day to day running. Their plan was to build a system for Auckland that in time could then be rolled out to other cities in New Zealand achieving economies of scale that would somehow “save” money. Snapper dug their heels in – they had a system in place and were gaining traction with micropayments but the NZTA wanted none of this. A deal was done with French giant Thales to provide the bulk of the backend systems for the ticketing solution and terminals for trains and ferries. Terminals for buses would be available from two vendors, with Snapper pledging their system could be made compatible meaning their terminals would stay.

I don’t need to write about the disaster that evolved over the next year or so. Thales struggled to deliver on time, the AIFS (Auckland Integrated Fare System)  which was key to the actual billing of journeys was a joke, and Snapper along with the other bus terminal vendor (who’s name completely escapes me right now) had trouble making their terminals compatible because they had a) nothing to test against because AIFS wasn’t built, and b) because the specifications kept changing. We all now know the outcome – Snapper pulled out of Auckland, Thales took over all bus terminals, and in 2012 HOP was launched with the cost blowing out to somewhere in the vicinity of $100 million. Since that day it’s either been plagued by problems or the best solution to ever hit the market  - depending solely on the side of the fence you want to sit on. Needless to say both parties believed in some very different things – Thales, NZTA and ARTA were very much about building closed, proprietary systems. Snapper were all about building open systems. It really is no wonder heads clashed with such differing views.

The involvement of the NZTA in Auckland was critical. Their involvement resulted in the creation of NITIS (national integrated ticketing interoperability standard) along with the concept of a centralised system for overall management, but with individual clearing houses for each city or town that wanted to jump onboard. This clearing house concept is quite important here –  unless fundamental changes are made to the clearing house model you will not be able to use an Auckland issued HOP card in any other region, or use (say) a Wellington HOP card in Auckland. The NZTA also don’t want their solution used for anything but public transport. Both approaches differ significantly to what Snapper believed in, with Snapper pushing it’s use in taxis and parking along with micropayments.

While it wasn’t clear to many at the time, the future for Snapper became a little less uncertain after HOP went live. With the NZTA committing to funding integrated ticketing outside Auckland on the condition integrated ticketing solutions were fully compliant with the NITIS specifications (something Snapper wasn’t) . Here in Wellington our slightly backwards GWRC still saw no need for integrated ticketing across all buses and trains and saw it as a solution looking for a problem, despite the fact the advantages of integrated ticketing across buses and trains are just … logical.

Move on to 2015 and integrated ticketing is finally on the cards at GWRC, with plans to have a solution in place towards the end of the decade. GWRC really only have one option for a solution, and that’s piggybacking on top of HOP. Deploying Snapper across the rest of the trains and buses in Wellington would cost a fraction of the cost of jumping onboard NZTA’s solution, and no doubt start raising all sorts of questions all over again about why HOP cost so much money not just on a solution, but a backwards solution.

10 years ago the concept of a piece of plastic to pay for public transport was a great one. It replaced cash and was reloadable. As we enter an era of smart devices and contactless payments in the form of Visa’s Paywave, Mastercard’s Paypass, Apple Pay and Samsung Pay the game has changed. Carrying around another piece of plastic to pay for a public transport journey feels backwards. Oyster card usage in London has plummeted this year with the adoption of Paywave, Paypass and Apple Pay to pay for all public transport journeys. Passengers no longer need to worry about carrying around an extra card and topping it up, and for tourists it makes using public transport incredibly simple.

The sad aspect of such advances is that the NZTA don’t believe in such things. Their concept of a closed model means HOP users are unlikely to see solutions such as this anytime soon. It’s highly likely you’ll see a mobile phone app but that’s hardly groundbreaking stuff – Snapper did that 3 years ago with their touch2pay solution which was the first of it’s kind in the world. Simply tagging your smartwatch against the reader to tag on and tag off  may be the way of the future in some countries, but New Zealand won’t be one.

All of this poses one big question. With all of the talent we have in New Zealand, how did we end up with the NZTA spending $100 million to build a solution based on French technology that doesn’t fully meet the needs of public transport users? The hardware isn’t the complex part, it’s the software that is. The clearing house and interchange model used by HOP is very inefficient, and issues such as online top-ups taking 3 days show the downsides of the current solutions.

Is it time to for Wellington to simply tell the NZTA were to stick HOP and build our own open standards solution for public transport ticketing?

Other related posts:
CCTV exposed. Why understanding network security is so important.
Anker make some of the best USB chargers and powerbanks available. Now you can get their products shipped directly to New Zealand
United Airlines pulls out of New Zealand for Southern Hemisphere Winter – AKL/SFO becomes seasonal.

Comment by aw, on 4-Dec-2015 11:23

As a daily user of Auckland's PT and a close follower of the integrated ticketing debacle up here, I have a few things to add.

JCOP wasn't part of Snapper from the start, it was introduced with a later version of their card (see this Snapper news post). It slowed things down.

These new cards (which all Auckland "Purple HOP" cards were) were slow to complete a transaction. I had (and still have) some older pre-JCOP Snapper cards from a 2009 Wellington trip, so I timed them. The older Snapper card took 400ms to tag on or off. The JCOP card was 600ms.

And the total time to beginning a tag transaction and being ready for the next one was 2000ms, meaning it took a minimum of 2 seconds per passenger to tag on and board. This adds up.

Thales-HOP is a little below 300ms and the reader is ready for the next card almost immediately - although strangely seems slower for two reasons - there's a 200ms delay between the transaction being complete - shown by the green LED coming on - and the terminal beeping. People wait for the beep which is an unnecessary slowdown. The second reason is latent - because Snapper readers weren't ready for the next card until the green LED went off again, and people learnt this behaviour, people wait for the same to happen for Thales-HOP.

Lastly, Snapper were quite aggressive to the point of being arrogant up here, especially from their behaviour once they lost the initial tender then pushed their systems onto (NZ Bus only) buses up here. The same ownership mistrust issues saw it getting no further, the other Auckland bus companies didn't want Infratil equipment on their buses just as you describe in Wellington.

The second terminal manufacturer was going to be Wayfarer/Parkeon, and from what I heard they were ready to go, but rollout was paused while dealing with Snapper, and when Snapper got the boot it became more economical to just get all the terminals from Thales (who weren't prepared to supply them to just a subset of bus companies, also due to economies of scale).
See also:

Comment by aw, on 4-Dec-2015 11:31

Oh and also see part two:

Comment by WyleECoyoteNZ, on 4-Dec-2015 11:57

While a integrated Wellington ticketing system would be nice, when I used snapper in the past, if I topped up $20.00, a 25c fee was deducted each time, leaving a balance on the snapper of $19.75.

I'm now a daily user of the Newlands\Mana bus service, where if I add $20.00 to my card, I get $20.00 worth of travel.

Ok, yes, this has to be done on-board the bus, and maybe can cause some delays, but that's a small price to pay.

Comment by knoydart, on 4-Dec-2015 14:53

What about the float? That is how Snapper bid for the Auckland system at "no cost" to the tax payer. Whoever controls the float, has a large pot of cash to sit on and if you were say a wholly owned subsidiary of your parent company who runs the largest bus company in the city, its a pretty good place to be (if you can get there). 

Comment by Hamish MacEwan, on 5-Dec-2015 08:46

Thanks once again Steve for a comprehensive explanation of a mundane activity which suffers from social and technical issues.Payments, well, that's a topic so big, incumbent, regulated and proprietary as to make disruption nigh on impossible.  It reminds me of Buckminster Fuller's dicta,  “You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.”But payment solutions inevitably appear to have to be built on the existing foundations and the payments tower of babel, and the associated fees and complexity hardly seem worth the candle where most of the margin is going to the incumbent oligopoly financial institutions and costs and risks are high.While there are new players, they appear obliged to treat with the dominant incumbents and once the State gets involved, particularly the intransigent and self-righteous "never wrong" NZTA, well, quagmire seems insufficient to describe the result.Solutions that serve incumbents, not the customer, are regularly like that, and in the payments space, it appears there are masters and supplicants.  It remains to be seen if any of the new entrants can elude the oppression of the established to overturn the model.  I have my doubts.Can a new model be built?  Given it's about devices and communication, in a mobile world with smart devices you'd think so.The financial network, is of course, much less accessible than the telecommunications network was, but eventually their grip and ruthless destruction of unruly competitors will fail.  Fortunately it always does.

Comment by Jaxson, on 5-Dec-2015 20:37

Interesting write up.


Comment by nigelj, on 9-Dec-2015 20:56

In addition to aw's posts...

AT Hop (as it's known now, ARTA hasn't existed since ~2010), is fairly reliable now, the issue with top ups potentially taking 3 days is easily explained, basically there is a sync window that starts at around 10pm and ends at X am, for depots & buses to sync the HOP systems onboard with pending actions (blacklists, Web topups, activations (for auto-topups), monthly pass purchases, and also in the other direction transaction history & auto-topup 'hits',  among I imagine other events...).   It's why there are large reminders that events performed online after 10pm may take extra time.  The 3 day window is just because it's entirely plausible that a bus is not returned to the depot overnight, or is parked too far away to connect to the sync network (I remember reading it was WiFi based).  I believe tag posts/gates at the ferry terminals/train stations bypass the 10pm sync because they are connected live.

In addition issues with charging have been pretty much put to bed (although when Auckland goes zonal rather than staged this might crop up again - except we already have a zone-esk system that works for Monthly passes so *crossed fingers*), the main issues that consumed a lot of time post-AT HOP launch were the Inner & Outer Link routes, circular routes with, you could say strange billing arrangements + the added issue that if the bus you were travelling on was going from point A to point E, but at point C there was a scheduled driver/shift/etc swap that required a system logoff/logon with a new trip number...  well lets just say, a lot of penalty fares got waived because of that until they figured out a better way.  I seem to recall Snapper didn't extremely well with that either.

In my opinion, it's not actually Thales that leaves the AT Hop system rather lackluster, it's some of the people at AT HQ.

Comment by muppet, on 11-Dec-2015 14:23

Sanpper's card was called "Hop".  I still have mine and use it when I visit Wellington.
The new Thales card was called "AT Hop", just to confuse the jesus out of people.

Comment by Tatou, on 12-Dec-2015 17:46

I top up Snapper with the NFC and Snapper App on my Samsung. Best top up App I have used and no charge for using Credit Card for topping up.


Comment by spoonboy, on 3-Jan-2016 03:41

Why these extra cards (Snappe, Thales) are required if there is a Visa payWave?
Why passengers should pay for the extra system and card if there is a standard and cheap way of doing contactless payments straight from their credit card???
Snapper has certainly failed making its cards something more than just a bus ticket. These days it's simply an expensive agent between a passenger's credit card and a transport company. So, we should get rid of it and reduce the costs.  I don't see why Snapper is called an "open system". It's like saying that Snapper is open just because the company is using Linux as a platform for running their proprietary application.  However, the alternative of spending money on building another ticketing system is even worse.

Comment by Nil Einne, on 6-Mar-2016 03:31

spoonboy: Visa paywave fees aren't exactly cheap although I'm not sure how they compare to AT Hop. More to the point neither payWave nor PayPass were really ready in NZ when Hop was being planned. Let's not forget it was supposed to be ready for the RWC. While this didn't happen in practice (perhaps typical NZ although as others have said, I think Snapper really delayed things), the RWC was when we got some of the first real marketing for payWave/PayPass. I.E. it was in its infacy. And even with the delays, AT Hop was still rolled out properly before paywave/PayPass were properly available i.e. on all major banks all credit and debit cards (actually I'm not sure if this is the case even now). In other words, it was foolish for AT/ARTA or NZTA to rely on the assumption third party credit card payment systems which they had little control over, let alone smart phone etc stuff would be ready. Heck as I understand it, TfL only rolled out contactless credit/debit card payments in 2014. Even if AT/NZTA had approached the credit card companies, NZ is small fry, realisticly we were never going to get much attention. In other words, yes we needed a system whatever was going to happen in the future. Perhaps we didn't give enough attention to how we can include these future advancements but then again with all the other problems particularly relating to Snapper perhaps it isn't surprising. The big problem is that electronic and integrated ticketing and fares came and is coming so so late to NZ and Auckland. (Let's not forget we still don't have integrated fares in Auckland. Sure I can use my AT Hop on all PT, but if I take a bus from the train or two buses or whatever all I get is a measly 25 cent or whatever discount and may get charged for a 4 stage and 1 stage trip when it should be 5 stages.) I mean when I came to NZ in 2002 they were talking about it (or perhaps it was a year or two later). Meanwhile in Malaysia I had used tag on/tag off contactless payment for buses in 1999. To be fair, although this commercial system was supposed to become the standard (with whatever dubious behind the scenes that is typical for Malaysia) a while before I left, it was I think 3 years or so after I left that you could finally used it for all public transport in Malaysia (including the LRT near my old house) and all tolls. Not untypical of Malaysia either but it still mean it was done about 6 years or so before NZ. To be fair, I also got a Ritchies contactless card in ~2005 (possibly it existed before then) although it wasn't a tag on an off system but a put the card on the machine and tell the driver (if you had cash rather than stored stages or both). As for using Snapper as an alternative, I'm with aw and nigelj here (and partly spoonboy). It's difficult to see Snapper as superior when experiences suggest they weren't, they were slow. And let's not forget whatever the problems in the specs, Snapper wasn't able to met them while others were. (Again while I don't claim to have a good enough memory to compare to 1999, I've spoken to people who've suggested Snapper in Auckland in 2011 was slower than touch'no'go in Malaysia of around the same time or a bit before.) There's no inherent reason giving all the money for a South Korean solution would be better than the giving all that money for a French solution. Neither were really NZ solutions and this is hardly surprising since as I said before, we're doing something over 10 years after even a developing country was already well on their way. So it was fairly dumb to reinvent the wheel. And at least some of that money was because of the apparent disaster of allowing Snapper to stay in the mix. Ultimately, it's impossible to prove whether things would have been a lot better, or pretty much the same if the parties involved had just told Snapper to bugger off once they lost. (And whatever personalities who didn't like NZ Bus, it does appear one apparent reason why they were allowed to stay was because one minister did like them.) P.S. As for the single supplier issue, yes in some ways it would be be good if AT Hop was open and we weren't beholden to any particular company. The problem is again we're doing this very late and very few others ever really did that. Plus we're so small fry that finding reliable suppliers interested is us is not easy Plus and this relates to both earlier points, no matter how good your spec is in theory, practice in oh so many areas shows that interoperability is often a problem.

Comment by Nil Einne, on 6-Mar-2016 03:34

I definitely had paragraphing, somehow it seemed to be lost ....

sbiddle's profile

Steve Biddle
New Zealand

I'm an engineer who loves building solutions to solve problems.

I also love sharing my views and analysis of the tech world on this blog, along with the odd story about aviation and the travel industry.

My interests and skillset include:

*VoIP (Voice over IP). I work with various brands of hardware and PBX's on a daily basis
  -Asterisk (incl PiaF, FreePBX, Elastix)

  -xDSL deployments

*Structured cabling
  -Home/office cabling
  -Phone & Data

*Computer networking
  -Mikrotik hardware
  -WAN/LAN solutions

*Wireless solutions
  -Motel/Hotel hotspot deployments
  -Outdoor wireless deployments, both small and large scale
  -Temporary wireless deployments
*CCTV solutions
  -Analogue and IP

I'm an #avgeek who loves to travel the world (preferably in seat 1A) and stay in nice hotels.

+My views do no represent my employer. I'm sure they'll be happy to give their own if you ask them.

You can contact me here or by email at